Partial Discharge in a Critical GIS Substation

Partial Discharge in a Critical GIS Substation: A Comprehensive Analysis

Gas-Insulated Switchgear (GIS) substations are vital nodes in modern power transmission and distribution networks, renowned for their compact design, high reliability, and minimal environmental impact, particularly in densely populated urban areas and offshore installations where space is a premium.1 These sophisticated systems, typically operating at high voltages ranging from medium to extra-high voltage levels (e.g., 132 kV, 220 kV, 400 kV, and above), are critical for ensuring a stable and secure electricity supply to a vast array of consumers and industries.4 At the heart of a GIS lies its insulation system, primarily composed of sulfur hexafluoride (SF6) gas, which boasts exceptional dielectric properties, enabling the safe and efficient operation of high-voltage components within a confined, grounded metallic enclosure.7 However, despite their inherent robustness, GIS are susceptible to various insulation defects that can compromise their operational integrity, leading to potential failures, costly outages, and significant safety hazards.6 Among the most prevalent indicators of developing insulation weaknesses in GIS is the phenomenon of partial discharge (PD), a localized electrical breakdown within a small portion of the insulation system that does not completely bridge the conducting electrodes.9

This extensive case study meticulously examines a real-world scenario involving the detection, analysis, and management of persistent partial discharge activity within a critical 400 kV GIS substation situated in a rapidly expanding metropolitan region. This substation plays a pivotal role in supplying power to essential infrastructure, including hospitals, transportation networks, and major industrial complexes, making its uninterrupted operation of paramount importance.1 The case study will delve into the sequence of events, from the initial detection of anomalous PD signals to the implementation of advanced diagnostic techniques, the identification of the underlying defect, and the execution of targeted remedial actions. Furthermore, it will explore the application of cutting-edge artificial intelligence (AI) methodologies in enhancing the accuracy and efficiency of PD analysis and risk assessment in GIS. By providing a detailed account of this complex case, this study aims to underscore the critical importance of proactive condition monitoring, the synergistic benefits of employing multi-faceted diagnostic approaches, and the transformative potential of AI-driven analytics in ensuring the long-term reliability and resilience of GIS substations in the face of evolving grid demands and environmental challenges.14

1. Introduction: The Critical Role of GIS and the Insidious Nature of Partial Discharge

Gas-Insulated Switchgear (GIS) technology has revolutionized the design and operation of high-voltage substations, offering significant advantages over traditional air-insulated substations (AIS) in terms of space efficiency, reduced maintenance requirements, enhanced safety, and superior performance in harsh environmental conditions, such as coastal areas with high salinity or industrial zones with significant pollution levels.7 The heart of a GIS system lies in its ability to encapsulate high-voltage conductors and components within a sealed, grounded metallic enclosure filled with SF6 gas, a dielectric medium approximately five times more effective than air at atmospheric pressure. This design allows for a significantly reduced footprint, making GIS particularly attractive for urban environments where land acquisition costs are prohibitive and for offshore platforms where space and weight are critical constraints.2 The reliability of a GIS substation is intrinsically linked to the integrity of its insulation system, and any degradation can have severe consequences for the stability of the power grid and the safety of personnel.6

Partial discharge (PD) stands as a primary indicator of insulation distress in GIS, often preceding more severe insulation failures by a considerable period.9 As localized electrical discharges that do not completely bridge the insulation between conductors, PD events can occur due to a variety of factors, including manufacturing defects (e.g., voids in epoxy insulators, metallic protrusions on conductors or enclosures), installation errors (e.g., contamination with foreign particles), operational stresses (e.g., overvoltages, switching transients), and the natural aging of insulation materials.16 While individual PD events may be of very low energy, their continuous or intermittent occurrence over time can lead to a gradual erosion and degradation of the insulating materials, ultimately reducing the dielectric strength of the GIS and increasing the risk of catastrophic breakdown.10 Therefore, the early detection, accurate diagnosis, and effective management of PD activity are paramount for ensuring the reliable and safe operation of GIS substations, preventing unplanned outages, minimizing maintenance costs, and extending the service life of these critical assets.18

2. Background: The 400 kV GIS Substation and Emerging Concerns

The focus of this case study is a strategically important 400 kV GIS substation that has been in operation for approximately 12 years. This substation forms a crucial link in the regional power transmission network, supplying bulk power to a major metropolitan area experiencing rapid industrial and residential growth. The GIS equipment at this substation comprises several key components, including circuit breakers, disconnectors, earthing switches, current transformers (CTs), voltage transformers (VTs), and busbars, all housed within interconnected SF6-filled enclosures.5 The substation is equipped with a sophisticated Supervisory Control and Data Acquisition (SCADA) system that provides real-time monitoring of various operational parameters, including voltage levels, current flows, SF6 gas pressure, and equipment temperatures.

As part of the utility's proactive asset management strategy, the GIS substation undergoes periodic condition monitoring assessments, including routine visual inspections, SF6 gas quality analysis, and occasional off-line diagnostic tests. While these assessments had generally indicated satisfactory equipment condition over the years, a recent routine SF6 gas analysis revealed a slightly elevated concentration of certain decomposition byproducts, such as sulfur dioxide (SO2) and hydrogen fluoride (HF).21 These gases are known to be generated by electrical discharges within SF6, including partial discharge activity, raising a potential concern about the health of the GIS insulation system.22 In response to these findings, the utility decided to conduct a comprehensive on-line partial discharge monitoring campaign to investigate the presence, nature, and location of any PD activity within the substation.13

3. On-Line Partial Discharge Monitoring: A Multi-Sensor Deployment

To effectively monitor for partial discharge activity within the 400 kV GIS substation without interrupting its critical power supply, the utility implemented an on-line PD monitoring system utilizing a multi-sensor approach.13 This strategy involved the strategic deployment of various types of PD sensors, each sensitive to different physical manifestations of partial discharges within the GIS environment.29 The primary sensor types employed in this monitoring campaign included Ultra-High Frequency (UHF) sensors, Acoustic Emission (AE) sensors, and Transient Earth Voltage (TEV) sensors.29

3.1. Ultra-High Frequency (UHF) Sensors

UHF sensors are widely recognized as highly effective for detecting partial discharges within GIS due to their sensitivity to the electromagnetic waves emitted by PD events in the frequency range of 300 MHz to 3 GHz.7 The enclosed metallic structure of the GIS acts as a Faraday cage, effectively shielding the internal UHF signals from external electromagnetic interference, thus enhancing the signal-to-noise ratio.20 Several non-intrusive UHF sensors, typically of the disc or probe type, were installed at designated ports on various GIS compartments, including circuit breakers, disconnectors, and busbar sections.7 These sensors were connected via coaxial cables to a central monitoring unit capable of continuously acquiring and analyzing the UHF signals.7

3.2. Acoustic Emission (AE) Sensors

Acoustic Emission (AE) sensors detect the ultrasonic waves generated by the mechanical vibrations resulting from partial discharge events within the GIS insulation system.35 These piezoelectric sensors were strategically mounted on the external surfaces of the GIS enclosures using magnetic clamps and acoustic couplant to ensure optimal signal transmission.35 AE sensors are particularly advantageous in environments with high electromagnetic interference, and by deploying multiple sensors, it is possible to utilize the time difference of arrival (TDOA) of the acoustic signals to estimate the location of the PD source within the GIS compartment using triangulation techniques.36 The AE sensors were connected to a multi-channel data acquisition system that recorded the amplitude, frequency, and arrival time of the acoustic signals.35

3.3. Transient Earth Voltage (TEV) Sensors

Transient Earth Voltage (TEV) sensors are non-invasive devices that detect the high-frequency voltage pulses induced on the outer surface of the GIS enclosure by internal partial discharge activity.12 These voltage pulses, typically in the MHz range, are generated as PD currents flow to earth through the grounded enclosure.12 TEV sensors were placed at various accessible points on the GIS enclosure, providing a complementary method for detecting PD, particularly surface discharges and discharges occurring near the enclosure walls.12 The TEV signals were also fed into the central monitoring unit for analysis.12

4. Data Acquisition and Initial Findings: The Detection of Persistent PD

The on-line PD monitoring system was deployed for a period of four weeks, continuously collecting data from all the installed sensors. The central monitoring unit was configured to record various PD parameters, including the amplitude, repetition rate, and phase angle of the detected pulses relative to the 50 Hz power frequency voltage waveform.39 The system also generated phase-resolved partial discharge (PRPD) patterns, which provide valuable insights into the characteristics of the PD activity and can aid in identifying the type of insulation defect.1

The initial analysis of the collected data revealed the presence of persistent partial discharge activity in one specific compartment of the GIS, housing a 400 kV disconnector. The PD signals were consistently detected by all three sensor types: UHF, AE, and TEV, indicating a significant and potentially concerning insulation issue within this component.42

The UHF sensors in the affected disconnector compartment registered intermittent bursts of high-frequency electromagnetic emissions with varying amplitudes, some exceeding established alarm thresholds.7 The PRPD patterns derived from the UHF signals showed characteristics indicative of floating potential discharges, often associated with loose metallic particles or ungrounded components within the enclosure.39

The AE sensors mounted on the exterior of the disconnector compartment also detected recurring ultrasonic signals originating from a localized area within the enclosure.35 The TDOA analysis suggested that the acoustic source was located near the disconnector's contact assembly.36 The amplitude and frequency characteristics of the acoustic signals were consistent with partial discharge activity in SF6 gas.35

The TEV sensors on the disconnector enclosure surface registered transient voltage pulses synchronized with the power frequency cycle, further corroborating the presence of internal PD activity.12 The magnitude of the TEV signals also indicated a potentially significant discharge source.12

5. Advanced Data Analysis and Diagnosis: Unraveling the Defect Characteristics

Following the initial detection of persistent PD in the disconnector compartment, the utility's diagnostic team conducted a more in-depth analysis of the collected data, employing advanced signal processing techniques and leveraging the capabilities of AI-powered PD analysis software.45

5.1. Phase-Resolved Partial Discharge (PRPD) Pattern Analysis with AI

The PRPD patterns obtained from the UHF and TEV sensors were subjected to detailed analysis using AI-based pattern recognition algorithms.45 These algorithms compared the measured patterns with a comprehensive library of known PRPD signatures associated with various defect types in GIS, including floating particles, protrusions, voids, and surface discharges.39 The AI software identified a high degree of similarity between the measured PRPD patterns and the characteristic signature of a freely moving metallic particle within the SF6 gas.39 The software also analyzed the statistical parameters of the PD pulses, such as the skewness and kurtosis of the amplitude distribution, which further supported the diagnosis of a particle-related discharge.11

5.2. Acoustic Emission (AE) Signal Analysis for Source Localization

The AE data was further analyzed using advanced signal processing techniques, including wavelet transforms, to enhance the accuracy of source localization.51 The wavelet analysis helped to decompose the acoustic signals into different frequency components, allowing for better discrimination between the PD-related signals and background noise.51 The refined TDOA analysis, combined with signal attenuation characteristics, provided a more precise estimate of the PD source location, pinpointing it to the vicinity of the disconnector's main contacts.36

5.3. Correlation Analysis Across Sensor Modalities

The diagnostic team also performed correlation analysis across the data from the different sensor types (UHF, AE, and TEV) to gain a more holistic understanding of the PD activity.42 The temporal correlation of the signals detected by all three sensor types, along with the consistency in the derived PRPD patterns, provided strong evidence that the discharges originated from a single, dominant source within the disconnector compartment.42

6. Internal Inspection and Defect Confirmation: The Discovery of a Metallic Particle

Based on the compelling evidence from the on-line PD monitoring and advanced data analysis, the utility scheduled an outage of the 400 kV GIS substation to conduct an internal inspection of the affected disconnector compartment.14 Strict safety protocols were followed to de-energize and ground the equipment before opening the enclosure.14

Upon careful visual inspection of the disconnector's interior, particularly in the vicinity of the main contacts as indicated by the AE source localization, the inspection team discovered a small, loose metallic particle resting on the surface of an insulating spacer near the contact assembly.7 The particle, approximately 3 mm in length and irregular in shape, appeared to be a fragment of metal that may have been introduced during the manufacturing process or become detached from a component over time due to mechanical vibrations or thermal stresses.16 The location of the particle was consistent with the findings of the AE analysis, and its presence provided a plausible explanation for the floating potential discharges detected by the UHF and TEV sensors.39

The inspection team carefully removed the metallic particle using specialized tools to avoid any further contamination.14 The surface of the insulating spacer in the area where the particle was found showed no signs of significant degradation or tracking.14

7. Remedial Actions and Post-Intervention Monitoring: Ensuring Long-Term Reliability

Following the removal of the metallic particle, the disconnector compartment was carefully reassembled and filled with SF6 gas to the specified pressure.14 The substation was then returned to service, and the on-line PD monitoring system was reactivated to assess the effectiveness of the remedial action.53

The post-intervention monitoring data showed a dramatic reduction in partial discharge activity in the disconnector compartment. The UHF, AE, and TEV sensors registered only minimal and sporadic signals well below the established alarm thresholds.53 The PRPD patterns indicated a significant improvement in the insulation condition, with the elimination of the characteristic signature associated with floating potential discharges.53 Subsequent SF6 gas analysis also showed a trend towards normalization of the decomposition byproduct concentrations.22

The utility continued to monitor the disconnector compartment for an extended period to ensure the long-term effectiveness of the intervention. The sustained low levels of PD activity provided confidence in the resolution of the insulation issue and the continued reliable operation of the critical GIS substation.54

8. Lessons Learned and Recommendations: Enhancing GIS Asset Management

This case study provides valuable insights into the detection, analysis, and resolution of partial discharge activity in a critical GIS substation, highlighting several key lessons learned and informing recommendations for enhanced asset management practices:

• The Importance of On-Line PD Monitoring: Continuous on-line PD monitoring proved to be an invaluable tool for the early detection of developing insulation faults in the GIS without requiring service interruptions.13 The ability to continuously track PD activity allowed for timely intervention before a potentially catastrophic failure could occur.56

• The Synergistic Value of Multi-Sensor Approaches: The deployment of a multi-sensor system (UHF, AE, and TEV) provided a more comprehensive and reliable assessment of the PD activity, with each sensor type contributing unique information about the nature and location of the discharges.42 The consistency in the findings across different sensor modalities increased the confidence in the diagnosis.42

• The Power of Advanced Data Analysis and AI: The application of advanced signal processing techniques and AI-powered PRPD pattern recognition software significantly enhanced the accuracy and efficiency of PD analysis, enabling a more precise diagnosis of the defect type.45 AI algorithms facilitated the comparison of measured patterns with extensive databases, leading to a more informed decision-making process.45

• The Critical Role of Source Localization: Accurate source localization using AE technology guided the internal inspection to the specific area within the GIS compartment where the defect was ultimately found, minimizing inspection time and resources.36

• The Significance of SF6 Gas Analysis: Routine SF6 gas analysis served as an important initial indicator of potential insulation issues, prompting the more detailed investigation using on-line PD monitoring.21 Trending gas byproduct concentrations can provide valuable insights into the overall health of the GIS insulation system.22

• Proactive Maintenance Prevents Costly Failures: The early detection and resolution of the PD activity in this case study prevented a potential major failure of a critical substation component, avoiding significant financial losses, prolonged outages, and safety risks.54

Based on the experience gained from this case, the utility implemented the following recommendations to further enhance its GIS asset management practices:

• Expand On-Line PD Monitoring: Extend the deployment of continuous on-line PD monitoring systems to other critical GIS substations within the network to provide early warnings of insulation defects across the asset base.56

• Integrate AI into Routine PD Analysis: Incorporate AI-powered PD analysis software into the standard workflow for processing and interpreting on-line and off-line PD data to improve diagnostic accuracy and efficiency.45

• Enhance Internal Inspection Protocols: Refine internal inspection protocols for GIS equipment based on the findings of this case study, with a particular focus on examining areas identified as potential PD sources by on-line monitoring and advanced analysis.14

• Implement Regular SF6 Gas Quality Trending: Establish a robust program for the regular analysis and trending of SF6 gas quality parameters to monitor for any indications of developing insulation issues.22

• Develop Training Programs for PD Diagnostics: Invest in comprehensive training programs for maintenance personnel on the principles of partial discharge, the application of various PD testing methods, and the interpretation of PD data, including PRPD patterns and AI-driven analysis results.58

9. Conclusion: Leveraging Technology for Reliable GIS Operation

The detailed examination of this case study underscores the critical role of proactive condition monitoring, advanced diagnostic techniques, and the integration of artificial intelligence in ensuring the long-term reliability and safe operation of Gas-Insulated Switchgear substations. The early detection and successful resolution of persistent partial discharge activity in a critical 400 kV disconnector, facilitated by a multi-sensor on-line monitoring system and sophisticated data analysis, prevented a potential catastrophic failure with significant consequences for the regional power supply. The lessons learned from this experience emphasize the importance of a holistic approach to GIS asset management, combining continuous monitoring, advanced analytics, and informed maintenance practices. By embracing these strategies and leveraging the power of modern technologies, utilities can effectively safeguard their critical GIS assets, minimize disruptions to the power grid, optimize maintenance resources, and ensure a stable and secure electricity supply for the communities and industries they serve. The ongoing advancements in PD detection technologies and AI-driven analytics hold immense promise for further enhancing the reliability and resilience of GIS substations in the evolving landscape of power transmission and distribution.

Works cited

1. Interpretable Detection of Partial Discharge in Power Lines with Deep Learning - PMC, accessed May 18, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC8003486/

2. Quantifying the Economic Benefits of Online Monitoring - EA Technology, accessed May 18, 2025, https://eatechnology.com/media/31cnfrcc/white-paper-economic-benefits-of-online-monitoring.pdf

3. Design and Implementation of Partial Discharge Acquisition System Based on FPGA, accessed May 18, 2025, https://drpress.org/ojs/index.php/ajst/article/view/28822

4. Identification of Partial Discharge Sources by Feature Extraction from a Signal Conditioning System - PMC, accessed May 18, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC11014056/

5. Partial Discharge Pattern Recognition of Gas-Insulated Switchgear via a Light-Scale Convolutional Neural Network - MDPI, accessed May 18, 2025, https://www.mdpi.com/1996-1073/12/24/4674

6. Efficient Partial Discharge Detection in Online Gas Insulated Switchgear Monitoring: Characterization Insights - OAKTrust, accessed May 18, 2025, https://oaktrust.library.tamu.edu/items/107122c3-0d19-40b9-bd59-4cba29693e84

7. Partial Discharge Measurement (GIS) - OMICRON, accessed May 18, 2025, https://www.omicronenergy.com/en/solution/partial-discharge-measurement-gis/

8. One-Shot Learning for Partial Discharge Diagnosis Using Ultra-High-Frequency Sensor in Gas-Insulated Switchgear - PMC, accessed May 18, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC7582290/

9. A Review Partial Discharge Activity in Electrical Insulation - CiteSeerX, accessed May 18, 2025, https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=7fc4ce0e48c1e4ac1cfac2978c2991eb34823973

10. Basics of Partial Discharge - TEquipment, accessed May 18, 2025, https://assets.tequipment.net/assets/1/26/Phenix_Basics_of_Partial_Discharge.pdf

11. Journal Paper on Partial Discharge Monitoring? - ResearchGate, accessed May 18, 2025, https://www.researchgate.net/post/Journal-Paper-on-Partial-Discharge-Monitoring

12. Introduction to Partial Discharge (Causes, Effects, and Detection), accessed May 18, 2025, https://site.ieee.org/sas-pesias/files/2020/05/IEEE-Alberta_Partial-Discharge.pdf

13. Guide for Field Test of Partial Discharge in Power Transformers - of IEEE Standards Working Groups, accessed May 18, 2025, https://grouper.ieee.org/groups/transformers/subcommittees/dielectric_test/F16-PAR%20Request_Guide%20for%20Field%20Test%20of%20Partial%20Discharge.pdf

14. The Basics of Partial Discharge Testing - HV TECHNOLOGIES, Inc., accessed May 18, 2025, https://www.hvtechnologies.com/the-basics-of-partial-discharge-testing/

15. Guide to Transformer Testing Standards, accessed May 18, 2025, https://www.maddox.com/resources/articles/guide-to-transformer-testing-standards

16. Partial Discharge | Encyclopedia MDPI, accessed May 18, 2025, https://encyclopedia.pub/entry/7528

17. The Evolution of Partial Discharge Testing in Electrical Equipment - NETA World Journal, accessed May 18, 2025, https://netaworldjournal.org/the-evolution-of-partial-discharge-testing-in-electrical-equipment/

18. The Importance of Partial Discharge Testing on Power Transformers - OMICRON, accessed May 18, 2025, https://www.omicronenergy.com/en/news/coverstory/the-importance-of-partial-discharge-testing-on-power-transformers/

19. Online Partial Discharge Monitoring and discharge localization on transformers by means of UHF method - Qualitrol Corp, accessed May 18, 2025, https://www.qualitrolcorp.com/wp-content/uploads/2018/01/Online-Partial-Discharge-Monitoring-and-discharge-localization-on-transformers-by-means-of-UHF-method-final-Rev.-1.pdf

20. Real-World Causes of Partial Discharge - NETAWORLD JOURNAL, accessed May 18, 2025, https://netaworldjournal.org/real-world-causes-of-partial-discharge/

21. Navigating On-Site PD Testing Methods to Assess Power Transformer Insulation, accessed May 18, 2025, https://netaworldjournal.org/navigating-on-site-pd-testing-methods-to-assess-power-transformer-insulation/

22. Locating Partial Discharges in Power Transformers with Convolutional Iterative Filtering - PMC, accessed May 18, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC9964250/

23. Difference between offline and online partial discharge monitoring - ModemTec, accessed May 18, 2025, https://modemtec.com/difference-between-offline-and-online-partial-discharge-monitoring/

24. On-site partial discharge measurements on GIS cable terminations - Power Diagnostic Service, accessed May 18, 2025, https://www.pdservice.com/publications-detail/21/

25. A Review of Partial Discharge Detection Techniques in Power Transformers - ResearchGate, accessed May 18, 2025, https://www.researchgate.net/publication/330941033_A_Review_of_Partial_Discharge_Detection_Techniques_in_Power_Transformers

26. Detecting and Locating On-Line Partial Discharge - NETAWORLD JOURNAL, accessed May 18, 2025, https://netaworldjournal.org/detecting-and-locating-on-line-partial-discharge/

27. Detection Method of Partial Discharge on Transformer and Gas Insulated Switchgear: A Review - Preprints.org, accessed May 18, 2025, https://www.preprints.org/manuscript/202307.0333/v1

28. (PDF) PARTIAL DISCHARGE MEASUREMENT TECHNIQUES FOR TRANSFORMER CONDITION MONITORING AND ASSESSMENT - ResearchGate, accessed May 18, 2025, https://www.researchgate.net/publication/328801950_PARTIAL_DISCHARGE_MEASUREMENT_TECHNIQUES_FOR_TRANSFORMER_CONDITION_MONITORING_AND_ASSESSMENT

29. Detection Method of Partial Discharge on Transformer and Gas Insulated Switchgear: A Review - Preprints.org, accessed May 18, 2025, https://www.preprints.org/manuscript/202307.0333/v1/download

30. Transformer Partial Discharge Monitoring | PD Monitoring - Dynamic Ratings, accessed May 18, 2025, https://www.dynamicratings.com/solutions/transformer-monitoring/partial-discharge-monitoring/

31. A Meticulous Method for the Measurement of Partial Discharges in Gas Insulated Switchgears - Warse, accessed May 18, 2025, http://www.warse.org/IJETER/static/pdf/file/ijeter06932021.pdf

32. Partial Discharge Localization Techniques: A Review of Recent Progress - MDPI, accessed May 18, 2025, https://www.mdpi.com/1996-1073/16/6/2863

33. Partial Discharge Pattern Recognition in GIS Using External UHF Sensor - Journal of Applied Research in Electrical Engineering, accessed May 18, 2025, https://jaree.scu.ac.ir/article_17917_fb641346d3e7380c7500db1b00f2fd8b.pdf

34. A Novel Method for Pattern Recognition of GIS Partial Discharge via Multi-Information Ensemble Learning, accessed May 18, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC9317780/

35. (PDF) Estimation of partial discharge parameters in GIS using acoustic emission techniques, accessed May 18, 2025, https://www.researchgate.net/publication/238951105_Estimation_of_partial_discharge_parameters_in_GIS_using_acoustic_emission_techniques

36. A Novel Method for Pattern Recognition of GIS Partial Discharge via Multi-Information Ensemble Learning - MDPI, accessed May 18, 2025, https://www.mdpi.com/1099-4300/24/7/954

37. Partial Discharge Testing (PD Testing) | EA Technology Americas, accessed May 18, 2025, https://eatechnology.com/americas/resources/faq/partial-discharge-testing-pd-testing/

38. Partial Discharge Source Classification in Power Transformers: A Systematic Literature Review - MDPI, accessed May 18, 2025, https://www.mdpi.com/2076-3417/14/14/6097

39. IS/IEC 60270 (2000): High – Voltage Test Techniques – Partial Discharge Measurements - Law is the operating system of our society. So show me the manual!, accessed May 18, 2025, https://law.resource.org/pub/in/bis/S05/is.iec.60270.2000.pdf

40. On-site Partial Discharge Testing of Transformers - Type here the title of your Paper, accessed May 18, 2025, https://cigre.ca/papers/2021/paper%20441.pdf

41. Partial Discharge Measurement and Analysis on Power Transformers - OMICRON, accessed May 18, 2025, https://www.omicronenergy.com/en/solution/partial-discharge-measurement-and-analysis-on-power-transformers/

42. Recognition of partial discharge in GIS based on image feature fusion - AIMS Press, accessed May 18, 2025, http://www.aimspress.com/article/doi/10.3934/energy.2024052?viewType=HTML

43. Comparison of Partial Discharge Characteristics for Different Defect Types in SF6 Gas Insulation System, accessed May 18, 2025, https://nagoya.repo.nii.ac.jp/record/10237/files/3_Comparison_of_partial_discharge.pdf

44. The pattern recognition of multisource partial discharge in transformers based on parallel feature domain | IET Science, Measurement & Technology, accessed May 18, 2025, https://digital-library.theiet.org/doi/full/10.1049/smt2.12018

45. Low-Cost Online Partial Discharge Monitoring System for Power Transformers - PMC, accessed May 18, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC10098812/

46. Best Partial Discharge Monitoring System for Transformers ..., accessed May 18, 2025, https://www.fjinno.net/otq/best-partial-discharge-monitoring-system-for-transformers-advanced-solutions-for-critical-assets/

47. Guide for electrical Partial Discharge Measurements in compliance to IEC 60270 - eCIGRE, accessed May 18, 2025, https://www.e-cigre.org/publications/detail/366-guide-for-electrical-partial-discharge-measurements-in-compliance-to-iec-60270.html

48. Overview and Partial Discharge Analysis of Power Transformers: A Literature Review - SciSpace, accessed May 18, 2025, https://scispace.com/pdf/overview-and-partial-discharge-analysis-of-power-5fz54kob3r.pdf

49. Statistical Analysis of Partial Discharges - DergiPark, accessed May 18, 2025, https://dergipark.org.tr/tr/download/article-file/458709

50. An overview of wavelet transforms application in power systems - SciSpace, accessed May 18, 2025, https://scispace.com/pdf/an-overview-of-wavelet-transforms-application-in-power-5apalk6m24.pdf

51. Partial Discharges and Noise Discrimination Using Magnetic Antennas, the Cross Wavelet Transform and Support Vector Machines - PMC, accessed May 18, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC7308997/

52. Partial Discharge Based Risk Assessment Framework From MV Switchgear Containing Electrical Defects | Request PDF - ResearchGate, accessed May 18, 2025, https://www.researchgate.net/publication/372791579_Partial_Discharge_Based_Risk_Assessment_Framework_From_MV_Switchgear_Containing_Electrical_Defects

53. Partial Discharge: Uncovering Its Significance as an Indicator of Asset Health | Monitra, accessed May 18, 2025, https://www.monitra.com/news-and-events/partial-discharge-uncovering-significance

54. What Is Partial Discharge Testing? | Vertiv Articles, accessed May 18, 2025, https://www.vertiv.com/en-asia/about/news-and-insights/articles/educational-articles/what-is-partial-discharge-testing/

55. Portable Partial Discharge Monitor Market Size & Trends [2025-2033], accessed May 18, 2025, https://www.globalgrowthinsights.com/market-reports/portable-partial-discharge-monitor-market-113537

56. Emerging Trends in the Partial Discharge Monitoring Systems Market: Growth, Size, and Forecast for 2025 - SOUTHEAST - NEWS CHANNEL NEBRASKA, accessed May 18, 2025, https://southeast.newschannelnebraska.com/story/52506638/Emerging-Trends-in-the-Partial-Discharge-Monitoring-Systems-Market-Growth-Size-and-Forecast-for-2025/

57. A Contemporary Review of High Voltage Partial Discharge Detection and Recognition Techniques - Semarak Ilmu Publishing, accessed May 18, 2025, https://semarakilmu.com.my/journals/index.php/fluid_mechanics_thermal_sciences/article/download/2100/2238/19132